Abstract 17056: Validation of the Adult Hypertrophic Cardiomyopathy Sudden Death Risk Calculator in a Pediatric Population
Background: Hypertrophic Cardiomyopathy (HCM) is a frequent cause of Sudden Cardiac Death (SCD) in children; however robust SCD risk predictors are lacking. Our study evaluated the adequacy of the European Society of Cardiology HCM SCD Risk calculator developed for adult HCM to predict 5-year risk of SCD in pediatric HCM.
Methods: Retrospective single centre review of children < 18 years with HCM (1978- 2015). To validate the adult calculator in our cohort, 5-year SCD risk score was calculated, and discrimination and accuracy assessments were performed. Three risk groups (low, medium, high) were constructed using the calculated risk score. Kaplan-Meier methods with log-rank tests were used to compare freedom from SCD; competing risk models were applied to analyze aborted SCD events. C-statistic was used to quantify the discriminatory ability of the risk scoring, the discrepancy between predicted and observed risk were also assessed.
Results: 291 children were eligible for inclusion. The median (IQR) age at first presentation was 6.5 (0.2 – 13.0 yrs) (64% male). Average z-scores for maximal LV wall thickness, and LA diameter, were 2.4 (± 2.6), 0.74 (± 2.04) respectively, and maximal LV outflow gradient was 22 (± 35) mmHg. 5.7% had family history of SCD, 11.7% had non-sustained ventricular tachycardia, 5.3% had unexplained syncope. 12 (4.1%) patients died of SCD, 2 (0.7%) with aborted SCD. 34 (11.7%) had ICD insertion. 144 patients with complete data were included in risk analysis. The c-statistic for discrimination analysis was 0.728 (95% CI [0.617, 0.838], p< 0.001), suggesting the adult SCD risk scoring was capable of differentiating between pediatric patients with low, medium and high SCD risk. However, on calibration analysis, the predicted risk in children was significantly lower than the observed risk at Year 5. Predicted versus observed risk was 0.25% vs 2.6% in low-medium risk group; 0.65% vs 8.7% in high risk group (p= 0.02).
Conclusion: The adult HCM SCD risk calculator can discriminate between pediatric patients with low, medium or high 5-year SCD risk; however the predicted risk underestimates the observed risk. The calibration factor in the adult calculator requires modification and/or additional discriminating factors prior to use its use in children.
Author Disclosures: K. Armstrong: None. J. Jeon: None. C. Fan: None. C. Manlhiot: None. J. Wilson: None. K. George: None. E. Stephenson: None. S. Mital: None.
- © 2016 by American Heart Association, Inc.